A case-control study of 230 patients and 230 "best friend or closest neighbor" controls is proposed to evaluate risk indicators of primary brain tumors (specifically histologically-confirmed astrocytic tumors: astrocytoma, glioma, glioblastoma multiforme, medulloblastoma). Despite the fact that brain tumors are the primary source of cancer mortality in youth, that a variety of brain tumor risk factors have been suggested, and that experimental data exist linking brain tumor development to specific chemical carcinogens and oncogenic viruses, no largepopulation, controlled study has been done to determine the risk factors of brain tumor. Through self-administered questionnaire and telephone interview, we propose to: 1. determine the relationship between demographic characteristics and risk of glial tumors. 2. evaluate the role of occupational and environmental exposures and risk. 3. evaluate the role of trauma. 4. evaluate the role of familial factors. 5. evaluate the role of infection as a risk factor. Although this study is analytic in design, it is hypothesis-generating in nature. The unique availability of a patient population which includes 100 new astrocytoma patients per year, and the active brain tumor research interests of the investigators, has created a situation under which a valid study of the epidemiology of these tumors may be completed in an efficient and inexpensive manner.